call admission control for voice/data integration in...
TRANSCRIPT
Call Admission Control for Voice/Data
Integration in Broadband Wireless Networks
Majid Ghaderi,Student Member, IEEE, and Raouf Boutaba,Senior Member, IEEE
School of Computer Science
University of Waterloo
Waterloo, ON N2L 3G1, Canada
Abstract
This paper addresses bandwidth allocation for an integrated voice/data broadband mobile wireless
network. Specifically, we propose a new admission control scheme called EFGC, which is an extension
of the well-known fractional guard channel scheme proposed for cellular networks supporting voice
traffic. The main idea is to use two acceptance ratios, one for voice calls and the other for data calls in
order to maintain the proportional service quality for voice and data traffic while guaranteeing a target
handoff failure probability for voice calls. We describe two variations of the proposed scheme: EFGC-
REST, a conservative approach which aims at preserving the proportional service quality by sacrificing
the bandwidth utilization; and EFGC-UTIL, a greedy approach which achieves higher bandwidth uti-
lization at the expense of increasing the handoff failure probability for voice calls. Extensive simulation
results show that our schemes satisfy the hard constraints on handoff failure probability and service
differentiation while maintaining a high bandwidth utilization.
Index Terms
Call admission control, voice/data integration, quality-of-service, broadband wireless networks.
1
Call Admission Control for Voice/Data
Integration in Broadband Wireless Networks
I. I NTRODUCTION
Emerging wireless technologies such as 3G and 4G will increase the cell capacity of wireless
cellular networks to several Mbps [1]. With this expansion of wireless bandwidth, the next
generations of mobile cellular networks are expected to support diverse applications such as
voice, data and multimedia with varying quality of service (QoS) and bandwidth requirements
[2]. Wireless links bandwidth is limited and is generally much smaller than that of wireline access
links. Therefore, for integrated voice/data mobile networks it is necessary to develop mechanisms
that can provide effective bandwidth management while satisfying the QoS requirements of both
types of traffic.
At call-level, two important quality of service parameters are thecall blocking probability(pb)
and thecall dropping probability(pd). An active mobile user in a cellular network may move
from one cell to another. The continuity of service to the mobile user in the new cell requires a
successful handoff from the previous cell to the new cell. The probability of a handoff failure is
calledhandoff failure probability(pf ). During the life of a call, a mobile user may cross several
cell boundaries and hence may require several successful handoffs. Failure to get a successful
handoff at any cell in the path forces the network to drop the call. The probability of such an
event is known as the call dropping probability.
Since dropping a call in progress has a more negative impact from the user perspective, handoff
calls are given higher priority than new calls in accessing the wireless resources. This preferential
treatment of handoffs increases the probability of blocking new calls and hence may degrade
the bandwidth utilization. The most popular approach to prioritize handoff calls over new calls
is by reserving a portion of available bandwidth in each cell to be used exclusively for handoffs.
Based on this idea, a number of call admission control (CAC) schemes have been proposed which
basically differ from each other in the way they calculate the reservation threshold [3]–[8].
Bandwidth allocation has been extensively studied in single-service (voice) wireless cellular
networks. Hong and Rappaport [3] are the first who systematically analyzed the famousguard
2
channel(GC) scheme, which is currently deployed in cellular networks supporting voice calls.
Ramjee et al. [9] have formally defined and categorized the admission control problem in
cellular networks. They showed that the guard channel scheme is optimal for minimizing a
linear objective function of call blocking and dropping probabilities while thefractional guard
channelscheme (FGC) [9] is optimal for minimizing call blocking probability subject to a hard
constraint on call dropping probability. Instead of explicit bandwidth reservation as in GC, the
FGC accepts new calls according to a randomization parameter called theacceptance ratio. One
advantage of FGC over GC is that it distributes the new accepted calls evenly over time which
leads to a more stable control [10].
Because of user mobility, it is impossible to describe the state of the system by using only
local information, unless we assume that the network is uniform and approximate the overall
state of the system by the state of a single cell in isolation. To include the global effect of
mobility, collaborative or distributed admission control schemes have been proposed [4]–[8],
[10], [11]. Information exchange among a cluster of neighboring cells is the approach adopted
by all distributed schemes.
In particular, Naghshineh and Schwartz [4] proposed a collaborative admission control known
as distributed call admission control(DCA). DCA periodically gathers information, namely
the number of active calls, from the adjacent cells to make, in combination with the local
information, the admission decision. It has been shown that DCA is not stable and violates
the required dropping probability as the load increases [10]. Levin et al. [5] proposed a more
sophisticated version of the original DCA based on theshadow clusterconcept, which uses
dynamic clusters for each user based on its mobility pattern instead of restricting itself (as
DCA) to direct neighbors only. A practical limitation of the shadow cluster scheme in addition
to its complexity and inherent overhead is that it requires a precise knowledge of the mobile’s
trajectory. Recently, Wu et al. [10] proposed a distributed scheme called SDCA based on the
classical fractional guard channel scheme which can precisely achieve the target call dropping
probability. A key feature of SDCA is the formulation of the time-dependent call dropping
probability which can be computed by the diffusion approximation of the channel occupancy.
One of the challenges in considering multi-services systems is that the already limited band-
width has to be shared among multiple traffics. Epstein and Schwartz [12] investigated complete
sharing, complete partitioning and hybrid reservation schemes for two classes of traffic, namely
3
narrow-band and wide-band traffic. In general, complete sharing strategy achieves the highest
bandwidth utilization [12].
Fixed and movable boundary schemes for bandwidth allocation in wireless networks were
studied by Wieselthier and Ephremides [13]. They concluded that movable boundary schemes
can achieve a better utilization than fixed boundary schemes for voice and data integration. Since
then, a number of papers have been published focusing on the performance of fixed and movable
boundary schemes given different assumptions and network configurations [14]–[20].
In particular, Haung et al. [18] proposed a bandwidth allocation scheme for voice/data inte-
gration based on the idea of movable boundaries (MB). In their scheme, bandwidth is divided
into two portions that can be dynamically adjusted to achieve the desired performance. However,
they completely neglected the prioritization of handoff calls over new calls and treated the two
identically. Yin et al. [19] proposed adual threshold reservation(DTR) scheme, which extends
the basic guard channel to use two thresholds, one for reserving channels for voice handoff, and
the other for limiting the data traffic into the network in order to preserve the voice performance.
An extended version of DTR which implements queueing for data calls (DTR-Q) was proposed
in [20]. In general, queueing of new/handoff calls, can further improve the performance of call
admission control [21]. The main limitation of DTR (DTR-Q) is that it is static, i.e., the two
reservation thresholds are fixed over time regardless of the state of the network. Interested readers
are referred to [22] for a comparison between DTR and MB schemes.
This paper introduces anextended fractional guard channel call admission mechanism(EFGC)
for integrated voice and data mobile cellular networks. EFGC maximizes the wireless bandwidth
utilization while satisfying a target call dropping probability and a relative voice/data service
differentiation. The main idea is to use two acceptance ratios for voice and data according to
the desired dropping probability of voice calls and the relative priority of voice calls over data
calls. Similar to [15]–[20], we assume that call dropping is not an important issue for data calls
and treat handoff and new data calls in the same way. We define the extended MINBLOCK [9]
problem as follows:
for a given cell capacity, maximize the bandwidth utilization subject to a hard constraint
on the voice call dropping probability and relative voice/data call blocking probability.
To the best of our knowledge, extending the basic fractional guard channel scheme to address
the extended MINBLOCK problem is a novel work. We follow an approach similar to the
4
admission control algorithm proposed by Wu et al. [10] to derive the acceptance ratios for voice
and data calls. The main features of EFGC are as follows:
1) EFGC is dynamic, therefore, adapts to a wide range of system parameters and traffic
conditions.
2) EFGC uses separate acceptance ratios for voice and data calls, therefore, it is very straight-
forward to enforce a relative or even strict service differentiation between voice and data
traffic.
3) EFGC is distributed and takes into consideration the information from direct neighboring
cells in making admission decisions.
4) The control mechanism is stochastic and periodical to reduce the overhead associated with
distributed control schemes. EFGC determines the appropriate control parameters such
as the control interval length in order to restrict the impact of the network to the direct
neighbors only.
The rest of the paper is organized as follows. Our system model, assumptions and notations are
described in section II. Section III is dedicated to the proposed admission control algorithm and
presents the analysis of the proposed algorithm in detail. In section IV, we discuss the estimation
of control parameters such as arrival rates, then we address the multiple handoffs problem and
control interval length. Extensive simulation results and their analysis are presented in section
V. Finally, section VI concludes this paper.
II. SYSTEM MODEL
As shown in Fig. 1, we consider a cellular system which carries both voice and data traffic.
We assume that wireless bandwidth is channelized where a channel can be a frequency, a time
slot or a code sequence. We define the basic bandwidth unit (BU) as the smallest amount of
bandwidth that can be allocated to a call, e.g., a channel. In this paper we focus on call-level
QoS parameters, therefore only call-level traffic dynamics are required for resource allocation and
admission control. More specifically, we assume that theeffective bandwidth[23]–[25] concept
is applied to each call. When employing this concept, an appropriate effective bandwidth is
assigned to each call and each call is treated as if it required this effective bandwidth throughout
the active period of the call. The feasibility of admitting a given set of connections may then
5
Cell i
Cell j
Cell k
HandoffVoice call
EFGC
New/HandoffData Call
New Voice Call
jir
kir
Fig. 1. Integration of voice and data at the base station of a cellular network.
be determined by ensuring that the sum of the effective bandwidths is less than or equal to the
total available bandwidth, i.e. the cell capacity.
We assume that each voice call requiresbv BUs and each data call requiresbd BUs for the
whole duration of the call. In the system under consideration, voice handoff calls have the highest
priority, then come new voice calls, and lastly the new and handoff data calls are considered. As
mentioned earlier, there is no prioritization of handoff data calls, and hence handoff data calls
are treated the same as new data calls.
The considered system is not required to be uniform. Each cell can experience a different
load, e.g., some cells can be over-utilized while others are under-utilized. Letk = {v, d} denote
the type of traffic, i.e.k = v for voice andk = d for data traffic. Below is the notation which
will be used throughout this paper.
• M : number of cells in the network
• Ai: the set of the adjacent cells of celli
• ci: the capacity of celli in terms of BUs
• Ri(t): bandwidth requirements (used capacity) in celli at time t in terms of BUs
• pf i: voice handoff failure probability in celli
• pQoS: target voice handoff failure probability to be guaranteed
• k: the service index for voice and data withk = v for voice andk = d for data
• λki : type-k new call arrival rate into celli
• 1/µk: type-k mean call duration
• 1/hk: type-k mean cell residency time
6
• T : length of the control period
• bk: bandwidth requirement of type-k calls in terms of BUs
• Nki (t): number of active type-k calls in cell i at time t
• rji: routing probability from cellj ∈ Ai to cell i
• bki : type-k call blocking probability in celli
• aki : type-k call acceptance ratio in celli
• αi: relative priority of voice traffic over data traffic in celli defined asαi = avi /a
di
• αQoS: target relative priority of voice traffic over data traffic to be guaranteed
• pkb : network-wide type-k call blocking probability
• pd: network-wide voice call dropping probability
• E[z]: the mean of random variablez
• V [z]: the variance of random variablez
• z: time-averaged value of random variablez
• z: measured (observed) value of random variablez
Let random variablestdkandtrk
denote the call duration (call holding time) and cell residency
time of a typical type-k call, respectively. Similar to [3], [9], [10], [12]–[22], we assume that
tdkand trk
are exponentially distributed. In the real world, the cell residence time distribution
may not be exponential but exponential distributions provide the mean value analysis, which
indicates the performance trend of the system. Furthermore, our proposed admission control
algorithm involves a periodic control where the length of the control period is set to much less
than the average cell residency time of a call to make the algorithm insensitive to this assumption.
A. Multiple Handoffs Probability
As mentioned earlier, in order to make the optimal admission decision, distributed schemes
regularly exchange some information with other cells in the network. Those cells involved in
the information exchange form acluster. Due to the intercell information exchange, base station
interconnection network incurs a high signalling overhead. Moreover, as the cluster size increases
the operational complexity of the control algorithm increases too. In particular, two major factors
affect the overhead and complexity of distributed CAC schemes; (1) frequency of information
exchange, and, (2) depth of information exchange, i.e. how many cells away information is
exchanged.
7
To reduce the overhead, distributed CAC schemes typically have a periodic structure in which
only at the beginning of control periods information exchange is triggered. Moreover, information
exchange is typically restricted to a cluster of neighboring cells. Note that, if the control interval
is too small then frequent communications increases the signalling overhead. On the other hand,
if the control period is too long then the state information stored locally may become stale.
Similarly, if the cluster is too small then the exchanged information will poorly reflects the state
of the network. On the other hand, a big cluster will lead to higher overhead. An efficient CAC
scheme must compromise between the frequency and depth of information exchange.
In this paper, we set the control interval in such a way that the probability of having multiple
handoffs in one control period becomes negligible. Therefore, we can effectively assume that
only those cells directly connected to a cell can influence the number of calls in that cell during
a control period. In a sense, we reduce the control interval in favor of a smaller cluster size. We
claim that using this technique, the signalling overhead will not increase, while the collected
information on the network status will be sufficiently accurate for the purpose of a stochastic
admission control. The reason is that: first, by decreasing the control interval, the probability
of multiple handoffs decays to zero exponentially (see section IV-C); second, a cluster shrinks
quadratically with decreasing the depth of information exchange (see below).
Without loss of generality, consider a symmetric network where each cell has exactlyA
neighbors. Consider celli and all the cells around it forming circular layers as shown in Fig. 2.
From celli, all the cells up to layern are accessible withn handoffs assuming that celli forms
layer 0. The number of cells reachable byn handoffs from celli denoted byM(n) is given by
M(n) = 1 +A+ · · ·+ nA
= 1 +1
2n(n + 1)A .
(1)
Therefore, by slightly reducing the control interval, we essentially achieve the same control
accuracy but with reduced signalling overhead. The problem of choosing the proper control
interval will be further addressed in section IV-C.
B. Handoff Failure and Call Dropping Probabilities
Although call dropping probability is more meaningful for mobile users and service providers,
calculating the handoff failure probability is more convenient. Therefore, our calculations in this
8
6
0
5
4
3
2
1
16
15
14 12
13
18
7
8
9
10
11
17
Layer 2
Layer 1
Layer 0
Fig. 2. A cellular system with 3 layers.
paper are based on the handoff failure probability,pf , which can be related to the call dropping
probability, pd, as follows (refer to [3] for more details):
pd =∞∑
H=0
(P vh )H(1− pf )
H−1pf =P v
hpf
1− P vh (1− pf )
, (2)
whereH is the number of possible handoffs during the life of a call, andP vh is the handoff
probability of a voice call before the call completes which can be computed by the following
equation:
P vh = Pr(tdv > trv)
=
∫ ∞
t=0
Pr(tdv > trv |trv) Pr(trv = t) dt
=
∫ ∞
t=0
hv exp(−µvt) exp(−hvt) dt =hv
µv + hv
(3)
therefore,
pf =pd
1− pd
(µv
hv
). (4)
It means that for a givenpd, the equivalentpf can be easily computed based on (4). Therefore,
in this paper it is assumed that a target handoff failure probabilitypQoS must be guaranteed for
voice calls. Notice that, exponential assumption is a necessary condition in deriving (3). Interested
readers are referred to [26], [27] for the handoff probability under general call duration and cell
residency distributions.
9
C. Time-Dependent Handoff and Stay Probabilities
We compute here some useful probabilities required for the rest of our discussion. LetP kh (t)
denote the probability that a type-k call hands off by timet and remains active untilt, given
that it has been active at time0. Also, letP ks (t) denote the probability that a type-k call remains
active in its home cell until timet, given that it has been active at time0. Then,
P kh (t) = Pr(trk
≤ t) Pr(tdk> t)
= (1− exp(−hkt)) exp(−µkt),(5)
and,
P ks (t) = Pr(trk
> t) Pr(tdk> t)
= exp(−(µk + hk)t) .(6)
These equations are valid as far as the memoryless property of call duration and cell residency
is satisfied. On average, for any call which arrives at timet′ ∈ (0, t], the average handoff and
stay probabilitiesP kh and P k
s are expressed as
P kh (t) =
1
t
∫ t
0
P kh (t− t′) dt′, (7)
P ks (t) =
1
t
∫ t
0
P ks (t− t′) dt′ . (8)
These integrals can be easily computed with respect to (5) and (6). Finally, letP kji(t) denote
the time-dependent handoff probability andP kji(t) denote the average time-dependent handoff
probability from cellj to cell i wherej ∈ Ai. It is obtained that
P vji(t) = P v
h (t) rji, (9)
P vji(t) = P v
h (t) rji, (10)
because voice handoff calls are always accepted if there is enough free bandwidth. Similarly,
P dji(t) = ad
i
[P d
h (t) rji
], (11)
P dji(t) = ad
i
[P d
h (t) rji
], (12)
because data calls are always subject to an acceptance ratioadi in cell i.
In next section, we will use the computed probabilities to find the maximum acceptance ratios
for voice and data calls with respect to the prespecified call dropping probability (pQoS) and
relative voice/data acceptance probability (αQoS).
10
III. A DMISSION CONTROL ALGORITHM
The proposed distributed algorithm, EFGC, consists of two components. The first component
is responsible for retrieving the required information from the neighboring cells and computing
the control parameters. Using the computed control parameters, the second component enforces
the admission control locally in each cell. The following sections describe these two components
in detail.
A. Distributed Control Algorithm
As mentioned earlier, to reduce the signalling overhead EFGC has a periodic structure. All the
information exchange and control parameter computations happen only once at the beginning of
each control period of lengthT . Several steps involved in EFGC distributed control are described
below:
1) At the beginning of a control period, each celli sends the following information to its
adjacent cells:
a) the number of active voice and data calls presented in the cell at the beginning of
the control period denoted byN vi (0) andNd
i (0), respectively.
b) the number of new voice calls,N vi , and new/handoff data calls,Nd
i , which were
admitted in the last control period.
2) Each celli receivesNkj (0) andNk
j from every adjacent cellj ∈ Ai.
3) Now, cell i uses the received information and those available locally to compute the
acceptance ratiosavi andad
i using the technique described in section III-C.
4) Finally, the computed acceptance ratiosavi andad
i are used to admit call requests into cell
i using the algorithm presented in section III-B.
Assume that all the cells have the same number of adjacent cells. LetA denote the number
of adjacent cells. Also, assume that all the required information can be sent from one cell to
another cell in one message. Then, the signalling overhead in terms of the number of exchanged
messages in one control period isA messages per cell.
B. Local Admission Control Algorithm
Let (m,n) denote the state of celli, where there arem voice calls andn data calls active
in the cell. DefineSi as the state space of celli governed by EFGC scheme. ThenSi can be
11
m+1,n
))(1(v
hv
m ++ µ
),( nmav
i
v
i
v
iλν +
),
()
(n
md i
ad i
d iλ
ν+
))(
1(
dh
dn
++
µ
m,n+1
m,n
Fig. 3. Extended fractional guard channel transition diagram.
expressed as
Si = {(m, n)|mbv + nbd ≤ ci} . (13)
Let aki (m, n) denote the acceptance ratio for type-k calls where the cell state is(m,n). Fig. 3
shows the state transition diagram of the EFGC scheme in celli for a typical state(m,n) ∈ Si. In
this figure,νki is the type-k handoff arrival rate into celli. At each state there are two acceptance
ratios for voice and data calls in such a way thatavi (m, n) = 0, if (m + 1, n) /∈ Si
adi (m, n) = 1
αiav
i (m,n), if (m, n) ∈ Si
(14)
There is a service differentiation (αi) between voice and data calls that governs the relation
between these two acceptance ratios. In this paper, we assume that this service differentiation is
specified apriori (αQoS) and EFGC should maintain it regardless of traffic conditions.
For an accurate control, the call blocking probability in each period is given by complementing
the acceptance ratio. Therefore, by averaging acceptance ratios over a number of control periods,
the call blocking probability is expressed asbki = 1−E[ak
i ]. Consequently, the average network-
wide call blocking probability for the considered network is given by
pkb =
∑Mj=i λ
ki b
ki∑M
j=i λki
. (15)
The pseudo-code for the local admission control in celli is given by the algorithm of Fig. 4.
In this algorithm,xk is a type-k call requestingbk BUs. The corresponding type-k acceptance
12
if (xk is a voice handoff call) then
if (Ri(t) + bv ≤ ci) then
accept call;else
reject call;end if
else /* new voice or new/handoff data call */if (Ri(t) + bk ≤ ci)&(rand(0, 1) < ak
i) then
accept call;else
reject call;end if
end if
Fig. 4. Local call admission control algorithm in celli.
ratio is aki . Also, rand(0, 1) is the standard random generator function. In the next section, we
will present a technique to compute the acceptance ratio vectorai = (avi , a
di ) in order to complete
this algorithm.
C. Computing Acceptance Ratios
It is assumed that by setting the control intervalT to an appropriate value, each call experiences
at most one handoff during a control period (see section IV-C for more detail). Therefore,
immediate neighbors of celli, i.e. Ai, are those which will affect the number of calls and
consequently the bandwidth usage in celli during a control period.
The proposed approach for computing the acceptance ratios includes the following steps:
1) Each celli uses the information received from its adjacents and the information available
locally to find the time-dependent mean and variance of the number of calls in the cell.
2) The computed mean and variance of the number of calls is used to find the mean and
variance of the bandwidth requirement process in the cell.
3) Having the mean and variance of the bandwidth requirement process, the actual time-
dependent bandwidth requirement process is approximated by a Gaussian distribution.
4) The tail of this Gaussian distribution is used to find the time-dependent handoff failure in
each celli.
5) Time-dependent handoff failure is averaged over control interval of lengthT to find an
average handoff failure probability for the whole period.
13
6) Using the computed handoff failure probability and the prespecified QoS constraints, i.e.
pQoS andαQoS, acceptance ratiosavi andad
i are computed.
The number of calls in celli at timet is affected by two factors: (1) the number of background
(existing) calls which are already in celli or its adjacent cells, and, (2) the number of new calls
which will arrive in cell i and its adjacent cells during the period(0, t] (0 < t ≤ T ). Let gki (t)
andnki (t) denote the number of background and new type-k calls in celli at timet, respectively.
A background type-k call in cell i will remain in cell i with probability P ks (t) or will handoff
to an adjacent cellj with probability P kij(t). A new type-k call which is admitted in celli at
time t′ ∈ (0, t] will stay in cell i with probability P ks (t) or will handoff to an adjacent cell
j with probability P kij(t). Therefore, the number of background calls which remain in celli
and the number of handoff calls which come into celli during the interval(0, t] are binomially
distributed. For a binomial distribution with parameterq, the variance is given byq(1−q). Using
this property it is obtained that
V ks (t) = P k
s (t) (1− P ks (t)), (16)
V kji(t) = P k
ji(t) (1− P kji(t)), (17)
V ks (t) = P k
s (t) (1− P ks (t)), (18)
V kji(t) = P k
ji(t) (1− P kji(t)) . (19)
where,V ks (t) andV k
ji(t) denote the time-dependent variance of stay and handoff processes, and,
V ks (t) and V k
ji(t) are their average counterparts, respectively.
The number of type-k calls in cell i is the summation of the number of background calls,
gki (t), and new calls,nk
i (t), of typek. Therefore, the mean number of type-k active calls in cell
i at time t is given by
E[Nki (t)] = E[gk
i (t)] + E[nki (t)], (20)
where,
E[gki (t)] = Nk
i (0)P ks (t) +
∑j∈Ai
Nkj (0)P k
ji(t), (21)
E[nki (t)] = (ak
i λki t)P
ks (t) +
∑j∈Ai
(akj λ
kj t)P
kji(t) . (22)
14
Similarly the variance is given by
V [Nki (t)] = V [gk
i (t)] + V [nki (t)], (23)
where,
V [gki (t)] = Nk
i (0)V ks (t) +
∑j∈Ai
Nkj (0)V k
ji(t), (24)
V [nki (t)] = (ak
i λki t)V
ks (t) +
∑j∈Ai
(akj λ
kj t)V
kji(t) . (25)
Note that given the arrival rateλki and the acceptance ratioak
i , the actual new call arrival rate
into cell i is given byλki a
ki (see section IV-B). Therefore, the expected number of call arrivals
during the interval(0, t] is given byaki λ
ki t.
Knowing the bandwidth requirement of each type of calls, the mean and variance of bandwidth
usage in celli at time t are given by
E[Ri(t)] = bvE[N vi (t)] + bdE[Nd
i (t)], (26)
V [Ri(t)] = b2vV [N v
i (t)] + b2dV [Nd
i (t)] . (27)
As we mentioned in section I, the cellular system considered in this paper is a broadband
wireless system with a capacity of several Mbps. In practice, 3G systems and beyond can be
considered as broadband wireless systems (for example a UMTS system can support up to 2
Mbps) [1], [2]. With this range of cell capacity it is reasonable to apply the central limit theorem
(this will be further discussed in section V-C). Thus, the bandwidth usage in each cell can be
approximated by a Gaussian distribution with meanE[Ri(t)] and varianceV [Ri(t)]. That is
Ri(t) ∼ G(E[Ri(t)], V [Ri(t)]
). (28)
Therefore, the original problem of maintaining a target handoff failure probabilitypQoS is
reduced to maintaining the bandwidth usage below the available capacityci at any point in
time t ∈ (0, T ]. Approximating the handoff failure probability by the overload probability, the
time-dependent handoff failure probabilityPf i(t) can be computed as follows:
Pf i(t) = Pr(Ri(t) > ci
), (29)
therefore,
Pf i(t) =1
2erfc
(ci − E[Ri(t)]√
2 V [Ri(t)]
), (30)
15
whereerfc(c) is the complementary error function defined as
erfc(c) =2√π
∫ ∞
c
e−t2 dt . (31)
Then the average handoff failure probability over a control period is given by
Pfi=
1
T
∫ T
0
Pf i(t) dt . (32)
Finally, to guarantee the target handoff failurepQoS, we should have
Pfi= pQoS . (33)
To solve (33) forai = (avi , a
di ) we need one more equation. This equation can be derived
with respect to the required service differentiation. Given the service conditionad = f(av), the
acceptance ratio vectorai = (avi , a
di ) can be found by numerically solving (33). Functionf is
such that0 ≤ f(avi ) ≤ 1 and f(0) = 0. In addition,f is uniformly increasing over[0, 1]. The
boundary condition is thatai ∈ [0, 1] × [0, 1], hence if Pfiis less thanpQoS even forav
i = 1
thenai is set to(1, f(1)). Similarly, if Pfiis greater thanpQoS even forav
i = 0, thenai is set
to (0, 0). In this paper, we only consider a constant service differentiation function denoted by
αi, whereadi = av
i /αi.
Finally, (33) can be solved using the bisection method [28]. Letξ denote the required numerical
precision. Then, the computational complexity of this technique isO(log 1/ξ), given that all
mathematical operations (including exponentiation and integration) can be performed inO(1).
IV. CONTROL PARAMETERS
In previous sections, we assumed that several parameters are known to the admission control
algorithm apriori. Among these parameters are the call arrival rates, mean call durations, mean
cell residency times and routing probabilities. In practice, all these parameters can be extracted
from measured field data using an estimation technique. Measurement and estimation units are
used for providing the required parameters to the admission control unit as shown in Fig. 5. One
useful estimation technique is presented in the following subsection.
16
Measurement
Unit
Estimation
Unit
Call Admission
Control Unit
System Data Control Parameters
Accept/R
eject
Call request
Arrival/Completion/Handoff
Fig. 5. Control unit diagram.
A. Parameter Estimation
A common technique for estimating the mean values from measurement data is theexpo-
nentially weighted moving average(EWMA) technique. Letz denote a control parameter to be
estimated, e.g., arrival rate, andz its measured (observed) value. A moving average estimator for
z at nth step is given byz(n) = (1− ε) z(n− 1) + εz(n− 1), whereε is a weighting factor that
should be specified with respect to the sampled observations ofz. In general, a small value of
ε can keep track of the changes more accurately, but is too sensitive to temporary fluctuations.
On the other hand, a large value ofε is more stable but could be too slow in adapting to real
traffic changes. By using this estimator, it can be verified thatE[z] = E[z]. However, EFGC
is independent of the estimation technique, and hence, it is possible to use more sophisticated
estimation techniques to achieve more accurate estimations (refer to [29], [30]).
We use the EWMA technique to compute the new call arrival rateλ into a cell of the network.
The only unknown parameter is the estimation coefficientε. As mentioned before, the accuracy
of the EWMA estimation depends onε. The goal is to chooseε in such a way to minimize the
estimation error. To measure the estimation error, we use the mean squared error (MSE) of the
estimations as expressed by
MSE =1
N
N∑i=1
(λ(i)− λ(i)
)2
(34)
whereN is the number of measurements. In our experiments we found thatε = 0.8 minimizes
the prediction error and achieves sufficiently accurate predictions.
17
B. Actual New Call Arrival Rate
In section III-C, we used productsakj λ
kj to compute the mean and variance of the number of
calls in celli (j ∈ Ai). Let us define theactual new call arrival rateinto cell j, denoted byλkj ,
as follows
λkj = ak
j λkj . (35)
In order to computeaki for the new control period we need to knowλk
j for every adjacent cell
j (j ∈ Ai). Similarly, cell j needs to knowλki in order to be able to computeak
j . Therefore,
every cell depends on its adjacents and vice versa. To break this dependency, instead of using
the actual value ofλkj , each celli estimates the actual new call arrival rates of its adjacents for
the new control period.
Let λkj (n) denote the actual new call arrival rate into cellj during thenth control period.
Also, let Nkj (n) denote the number of new calls that were accepted in cellj during thenth
control period. Similar to [4], [10], an estimator forλkj is expressed as
λkj (n + 1) = (1− ε)
Nkj (n)
T+ ελk
j (n), (36)
where,λkj (n + 1) is the actual new call arrival rate into cellj at the beginning of the(n + 1)th
control period. Note thatλkj (n) is known at the beginning of the(n+1)th control period. In our
simulations we found thatε = 0.3 leads to a good estimation of the actual new call arrival rate.
C. Control Interval
The idea behind at-most-one handoff assumption is that by setting control interval appropri-
ately, the undesired multiple handoffs during a control period can be avoided. As discussed in
section III, this minimizes the signalling overhead and operational complexity of EFGC. In this
section, we address the control interval selection problem.
Consider a symmetric network where each cell has exactlyA neighbors, and the probability
of handoff to every neighbor is the same. Then, the routing probabilityrij from cell i to cell j
is given by
rij =
1/A, j ∈ Ai,
0, j /∈ Ai .(37)
18
Let q(n) denote the probability that an active call experiencesn handoffs during time intervalT .
Also, let qij(n) denote the probability that a call originally in celli moves to cellj over a path
consisting ofn handoffs during time intervalT . Define δ as the multiple handoffs probability
from cell i to cell j. We then can write
δ =∞∑
n=2
qij(n) . (38)
Our goal is to find a relation betweenT andδ in order to be able to controlδ by controllingT .
For an effective control (pf in the range of10−4 to 10−2) we can assume thatpf is effectively
zero. Similarly, if δ ≈ pf for a givenT , we can assume that the multiple handoffs probability
is zero. Since cell residency is exponential, the number of handoffs a call experiences during
an interval is Poisson distributed with meanhT , given that the call is active during the whole
interval. Therefore, it is obtained that
q(n) =(hT )n
n!e−(h+µ)T . (39)
In order to computeqij(n) based on (39), we need to find the probability of moving from cell
i to cell j by n handoffs. LetLij(n) denote the number of paths consisting ofn handoffs from
i to j, then
qij(n) =Lij(n)
Anq(n) . (40)
Consider the network depicted in Fig. 2. LetT = 20 s, 1/µ = 180 s, 1/h = 100 s andA = 6.
Table (I) shows the maximum probability of multiple handoffs from any cellj to cell 0, Pj0(n),
based on the number of handoffs,n. For eachn, we have also determined which layer has the
maximum paths to cell0. Interestingly, cell0 has the most paths to itself through other cells.
We have also illustrated in Fig. 6 the impact of the control intervalT on the multiple handoffs
probability δ for the same set of parameters.
Consider celli and all the cells around it forming circular layers. From celli, all the cells up
to layer n are accessible withn handoffs assuming that celli forms layer 0. It can be shown
that
Lij(n) ≤ An−1, n ≥ 1 (41)
because forn ≥ 1, at each level there are at leastA cells which have the same number of paths
to the destination celli. Therefore
qij(n) ≤ 1
A(hT )n
n!e−(h+µ)T , n ≥ 1 . (42)
19
TABLE I
MULTIPLE HANDOFFS PROBABILITY FORT = 20 s.
n Layer max{Lj0(n)} max{Pj0(n)}
0 0 1 0.73263
1 0 1 0.02442
2 0 6 0.00244
3 1 15 0.00007
4 0 90 0.00000
5 0 360 0.00000
5 10 15 20 25 30 35 40 45 5010
−4
10−3
10−2
Control interval (T)
Mul
tiple
han
doff
prob
abili
ty (
δ)
Fig. 6. Effect ofT on multiple handoffs probability.
Using (38) and (42), it is obtained that
δ ≤∞∑
n=2
1
A(hT )n
n!e−(h+µ)T
=ehT − hT − 1
Ae(h+µ)T.
(43)
Using the Taylor expansion of exponential terms forδ � 1A( h
µ+h), it is obtained that
T ≤ Aδ(µ + h) + h√
2Aδ
Aδ(µ + h)2 − h2, (44)
which finally leads to the following simple relation
T ≈√
2Aδ
h. (45)
20
V. SIMULATION RESULTS
A. Greedy EFGC
The basic EFGC introduced in section III may seem to be too conservative about accepting
data calls. We refer to this restrictive version of EFGC by EFGC-REST (or simply REST).
REST is a conservative approach which aims at satisfying the specified priority functionf over
time. In other words, REST always uses the acceptance ratioai = (avi , f(av
i )) regardless of the
congestion situation to impose an exact priority function.
It is observed that in some states of the system it is possible to increase the acceptance ratio
of data calls beyond the limit returned by the service differentiation function. For example when
the network is not congested (at light traffic loads), we found that by increasing the priority of
data traffic the overall utilization of the wireless bandwidth is increased while the handoff failure
remains almost untouched. This relaxed version is called EFGC-UTIL (or simply UTIL) due to
its greedy behavior in maximizing the bandwidth utilization. To find the data acceptance ratio
in cell i, UTIL follows the following steps:
1) Find avi using (33),
2) If (avi == 1) then find the maximum value ofad
i ∈ [f(1), 1] which satisfies (33),
It is worth noting that the computational complexity of EFGC-UTIL is the same as EFGC-REST,
i.e. O(log 1/ξ).
B. Simulation Parameters
Simulations were performed on a two-dimensional cellular system consisting of 19 hexagonal
cells (see Fig. 2). Opposite sides wrap-around to eliminate the finite size effect. It is assumed that
mobile users move along the cell areas according to a uniform routing pattern. In other words, all
neighboring cells have the same chance to be chosen by a call for handoff, i.e.rji = 1/6. For ease
of illustrating the results, the simulated system is uniform, i.e. input load is the same for every
cell, although EFGC as well as the simulation program are designed to handle the nonuniform
case as well. Therefore, unless explicitly specified, the subscripti is omitted hereafter.
The common parameters used in the simulation are as follows. All the cells have the same
capacity c = 5 Mbps, which is equal to 160 BU assuming each BU is equal to32 Kbps
(encoded voice using ADPCM requires 32 Kbps). Target handoff failure probability for voice
21
TABLE II
VOICE/DATA SERVICE PARAMETERS.
Type Priority 1/µ (s) 1/h (s) BU Load
voice 1 180 100 1 60%
data 0.5 1000 800 2 40%
calls ispQoS = 0.01 andT = 20 s. We use normalized load in simulations which is simply the
total arrival load per BU. Letρ denote the total normalized arrival load into a cell, then
ρ =1
c
(ρv + ρd
), (46)
where,ρv andρd are, respectively, voice and data load given by
ρv = bvλv/µv, (47)
ρd = bdλd/µd . (48)
For each load, simulations were done by averaging over 8 samples, each for 10 hours of
simulation time. Load distribution between voice and data traffic is fixed over time. At any
load, 60% of the load is due to voice calls and the remaining40% is composed of data calls.
Table II summarizes service and traffic parameters for both traffic types. In this table,priority
refers to the relative priority (service differentiation) of voice and data calls. It means that new
voice calls have higher priority than data calls for the admission control algorithm. In particular,
the probability of accepting a new voice call is at least twice the probability of accepting a data
call (new/handoff) at any time and any load. Equivalently, this is achieved by settingαQoS = 2.
As mentioned earlier, this relative priority can be any service differentiation function. In our
simulations, for the sake of simplicity we have chosen a constant service differentiation function.
We have also implemented the DTR scheme introduced in section I for comparison purposes.
Since DTR is designed for a static traffic pattern, the handoff failure probability increases rapidly
with the network load when the guard channels for handoff are few, but remains too low when
the guard channels are many. Here, we choose the two thresholds in such a way that DTR
achieves its objectives when the network starts to get overloaded. Hence, the voice threshold is
set to 155 BUs and the data threshold is set to 151 BUs. Using these thresholds at load 2,pf
andα = av/ad were found to be 0.01 and 2, respectively.
22
C. Gaussian Approximation
When the network is not congested and each cell has only a few active calls, it is clear that
Gaussian approximation is not good. However, at light loads the admission algorithm does not
require a high precision estimation of the load since there is no congestion in the network. As
the load increases the number of active calls in each cell increases rapidly until no more calls
can be accepted. Due to the high capacity of broadband systems, it is expected to have enough
active calls in each cell so that central limit theorem can be applied.
Other researchers have also successfully applied Gaussian approximation for similar purposes.
Naghshineh and Schwartz [4] and Epstein and Schwartz [7] used the same kind of approximation.
The main difference is that we extend their single point approximation at the end of the control
period to a time dependent approximation over the whole control period. The authors of [10] also
realized that for large system sizes, as is the case in this paper, the cell occupancy distribution
evolves into a Gaussian distribution. Investigating the tail behavior of the bandwidth usage
distribution is beyond the scope of this paper, instead we rely on the results from other researchers
[4], [7], [10], [23].
D. Results and Analysis
1) Effect of arrival load: The first set of simulation results show the main performance
parameters of EFGC. Fig. 7 shows the handoff failure probability for the three schemes for a
wide range of loads. Both UTIL and REST maintain a constant failure probability independent
of the load. For DTR, it grows very rapidly with the load (which was expected). With light
loads (load< 2), DTR and REST have almost the same handoff failure probability while UTIL
has slightly higher handoff failure probability. But with high loads (load> 2), UTIL and REST
converge to exactly the same handoff failure probability while DTR has much higher handoff
failure probability. Fig. 9(b) shows that, although REST has better failure probability in light
loads, this is accomplished at the expense of the data call blocking probability. However, even
in this region (load< 2), UTIL satisfies the target handoff failure probabilitypQoS.
One of the objectives of EFGC is to maintain the relative service priority between voice
and data calls. In our simulations, this relative priority is fixed and indicates that the acceptance
probability of new voice calls should be twice the acceptance probability of new data calls. Fig. 8
gives the service differentiationα = av/ad for different loads. It shows that EFGC maintains an
23
0 1 2 3 4 5 60.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
Han
doff
Fai
lure
Pro
babi
lity
Normalized Load
DTREFGC−UTILEFGC−REST
Fig. 7. Voice handoff failure probability.
almost constant service priority between the two types of traffic. More precisely, REST preserves
α = 2 for the whole range of loads while UTIL hasα = 1 in light loads andα = 2 in high loads
as expected. This can be explained by the fact that in light loads UTIL accepts data calls as long
as there is free bandwidth (without violating the target voice handoff failure probability). As
the load increases, service priority of DTR increases rapidly. Fig. 9(b) shows that at high loads
almost no data calls are accepted. In other words, DTR is not fair and leads to starvation of
data traffic. It is worth mentioning that, although in this simulation the service differentiation is
fixed, the EFGC can satisfy more complex priority disciplines such as state dependent priorities.
Fig. 9 shows the new voice and new/handoff data call acceptance probabilities respectively.
Again for high loads, UTIL and REST converge to the same result but the difference in their
performance at light loads is significant. For data traffic at light loads the acceptance probability
of UTIL is almost twice that of REST. This explains why the utilization of UTIL is superior to
REST. It can be seen that DTR has slightly higher acceptance probability for voice but much
lower acceptance probability for data in comparison to UTIL and REST.
Finally, Fig. 10 shows the wireless bandwidth utilization under the three bandwidth allocation
mechanisms. Although DTR performs poorly in terms of handoff failure probability and service
priority, its utilization is slightly better than EFGC. Interestingly, UTIL has exactly the same
utilization level as DTR at light loads but higher than that of REST. In this simulation, voice
24
0 1 2 3 4 5 60
2
4
6
8
10
12
14
16
18
20
Ser
vice
Diff
eren
tiatio
n
Normalized Load
DTREFGC−UTILEFGC−REST
Fig. 8. Voice/Data relative acceptance probability (α).
0 1 2 3 4 5 60.0
0.2
0.4
0.6
0.8
1.0
Acc
epta
nce
Pro
babi
lity
Normalized Load
DTREFGC−UTILEFGC−REST
(a) New voice calls acceptance probability.
0 1 2 3 4 5 60.0
0.2
0.4
0.6
0.8
1.0A
ccep
tanc
e P
roba
bilit
y
Normalized Load
DTREFGC−UTILEFGC−REST
(b) New/handoff data calls acceptance probability.
Fig. 9. Acceptance probability of voice and data.
traffic constitutes the larger portion of the total load. As the percentage of data traffic increases,
the utilization of DTR is expected to drop. This will be investigated next.
2) Effect of load sharing:In previous simulations, the load sharing factorβ(β > 0) is set to
1.5, where
β =arriving data traffic load (ρv)arriving voice traffic load (ρd)
. (49)
Due to the priority of voice calls over data calls, varyingβ will affect the behavior of EFGC.
As shown in Fig. 11, EFGC is insensitive to the load sharing factor. In these plots, theX axis
25
0 1 2 3 4 5 60.0
0.2
0.4
0.6
0.8
1.0
Ban
dwid
th U
tiliz
atio
n
Normalized Load
DTREFGC−UTILEFGC−REST
Fig. 10. Wireless bandwidth utilization.
indicates the load sharing factorβ. It is assumed that most of the traffic is composed of voice
calls, henceβ varies between 0.5 and 5.
For this set of simulations, normalized arrival load is set to1.5 Erlang and voice priority is
set to 2 (α = 2). As expected, DTR is not able to adjust to changes in load shares although the
total load is fixed. Interestingly asβ increases, EFGC-UTIL and EFGC-REST converge to the
same value for handoff failure probability. The reason is that by increasingβ, voice traffic will
dominate data traffic. Therefore, a larger portion of the available bandwidth is allocated to voice
traffic in such a way that there is no extra free bandwidth to be assigned to data traffic (more
than their guaranteed share).
The primary goal of the following set of simulations is to show the stability of EFGC under
various QoS requirements (pQoS and αQoS) and the insensitivity of EFGC to the exponential
assumption we made about the cell residency time.
3) Effect of voice priority:Fig. 12 shows the effect of changing the relative priority of data
calls and voice calls. In this set of plots, theX axis indicates the quantity1/α, where
1/α =data calls acceptance probability (ad)voice calls acceptance probability (av)
. (50)
In the simulations, the total arrival load is set to1.5 Erlang which consists of60% voice traffic
and40% data traffic (i.e. a load sharing factor of 1.5). It is found that regardless ofα, EFGC is
able to satisfy the targetαQoS while providing the desired service differentiation. The straight
26
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.000
0.005
0.010
0.015
0.020
0.025H
ando
ff F
ailu
re P
roba
bilit
y
Voice Load / Data Load
DTREFGC−UTILEFGC−REST
(a) Voice handoff failure probability.
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00.0
0.5
1.0
1.5
2.0
2.5
3.0
Ser
vice
Diff
eren
tiatio
n
Voice Load / Data Load
DTREFGC−UTILEFGC−REST
(b) Relative acceptance probability (α).
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0
0.2
0.4
0.6
0.8
1.0
Cha
nnel
Util
izat
ion
Voice Load / Data Load
DTREFGC−UTILEFGC−REST
(c) Wireless bandwidth utilization.
Fig. 11. Effect of load sharing (β).
lines in Fig. 12(b) indicate that any value of service differentiation can be strictly guaranteed
with EFGC.
As indicated in these figures, UTIL and REST converge to the same control policy asα tends
towards 1. This was expected because the two schemes differ from each other with respect toα.
In this case, available resources are completely shared among voice and data traffic and channel
utilization is maximized. However, for large values ofα (small values of1/α), UTIL has a
superior performance over REST. For example, atα = 1/0.2, UTIL has 4% better utilization.
27
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 110
−4
10−3
10−2
Han
doff
Fai
lure
Pro
babi
lity
Data/Voice Relative Priority
EFGC−UTILEFGC−REST
(a) Voice handoff failure probability.
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Dat
a/V
oice
Rel
ativ
e A
ccep
tanc
e P
roba
bilit
y
Data/Voice Relative Priority
EFGC−UTILEFGC−REST
(b) Relative acceptance probability (1/α).
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10.80
0.82
0.84
0.86
0.88
0.90
0.92
0.94
0.96
0.98
1.00
Cha
nnel
Util
izat
ion
Data/Voice Relative Priority
EFGC−UTILEFGC−REST
(c) Wireless bandwidth utilization.
Fig. 12. Effect of voice priority.
4) Effect of handoff failure probability (QoS):In cellular systems, the targetpQoS is typi-
cally set to1%. To show the adaptiveness of EFGC, simulations were performed forpQoS =
[0.2%, 1%, 5%]. Notice thatpQoS = 0.2% is an extremely low handoff failure probability. As
shown in Fig. 13, handoff failure and service differentiation are fully satisfied regardless of the
target QoS requirements. In particular, Fig. 13(a) shows the stability of EFGC under different
target dropping requirements.
5) Effect of non-exponential cell residency:The first part of our analysis, which gives the
equations describing the mean and variance of channel occupancy (i.e., number of busy channels
in a cell), is based on the exponential cell residency time assumption. This assumption may not be
28
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.010
−6
10−5
10−4
10−3
10−2
10−1
Han
doff
Fai
lure
Pro
babi
lity
Normalized Load
EFGC−UTIL: 5%EFGC−REST: 5%EFGC−UTIL: 1%EFGC−REST: 1%EFGC−UTIL: 0.2%EFGC−REST: 0.2%
(a) Voice handoff failure probability.
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00.5
1.0
1.5
2.0
2.5
Ser
vice
Diff
eren
tiatio
n
Normalized Load
EFGC−UTIL: 5%EFGC−REST: 5%EFGC−UTIL: 1%EFGC−REST: 1%EFGC−UTIL: 0.2%EFGC−REST: 0.2%
(b) Relative acceptance probability (α).
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cha
nnel
Util
izat
ion
Normalized Load
EFGC−UTIL: 5%EFGC−REST: 5%EFGC−UTIL: 1%EFGC−REST: 1%EFGC−UTIL: 0.2%EFGC−REST: 0.2%
(c) Wireless bandwidth utilization.
Fig. 13. Effect of handoff failure probability (QoS).
correct in practice and needs more careful investigation as pointed out in [31]–[33] and references
there in. Although exponential distributions are not accurate in practice but the models based
on the exponential assumption are tractable and do provide mean value analysis which indicates
the system performance trend.
Using real measurements, Jedrzycki and Leung [31] showed that a lognormal distribution is
a more accurate model for cell residency time. We now compare the results obtained under
exponential distribution with those obtained under more realistic lognormal distribution. The
mean and variance of both distributions are the same (refer to Table (II)). Fig. 14 shows that
29
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.010
−5
10−4
10−3
10−2
10−1
Han
doff
Fai
lure
Pro
babi
lity
Normalized Load
EFGC−UTIL: exponentialEFGC−REST: exponentialEFGC−UTIL: lognormal EFGC−REST:lognormal
(a) Voice handoff failure probability.
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00.5
1.0
1.5
2.0
2.5
Ser
vice
Diff
eren
tiatio
n
Normalized Load
EFGC−UTIL: exponentialEFGC−REST: exponentialEFGC−UTIL: lognormalEFGC−REST: lognormal
(b) Relative acceptance probability (α).
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cha
nnel
Util
izat
ion
Normalized Load
EFGC−UTIL: exponentialEFGC−REST: exponentialEFGC−UTIL: lognormalEFGC−REST: lognormal
(c) Wireless bandwidth utilization.
Fig. 14. Effect of non-exponential cell residency.
the exponential cell residency achieves sufficiently accurate control. In other words, the control
algorithm is rather insensitive to this assumption due to its periodic control in which the length
of the control interval is much less than the average cell residency time.
VI. CONCLUSION
In this paper, we proposed a new admission control algorithm for voice/data integration in
broadband wireless networks. Our algorithm is a natural extension of the well-known fractional
guard channel proposed for voice cellular systems. EFGC always achieves the predetermined call
dropping probability for voice calls while keeping the relative blocking probability of voice and
30
data calls within a target threshold. We then described two versions of the EFGC, namely EFGC-
UTIL and EFGC-REST. EFGC-UTIL follows a greedy approach to maximize the bandwidth
utilization while EFGC-REST maintains the relative service priority. Both versions converged
to the same result for high traffic loads. The major advantage of EFGC is its insensitivity to
network traffic load. The dropping probability of voice calls and relative blocking probability of
voice and data calls is maintained at a stable level over a wide range of traffic loads. From the
simulation results, we conclude that EFGC-UTIL is a better candidate for integrated voice/data
cellular networks.
We are currently investigating the case of multiple classes of traffic where each class has its
own QoS requirements in terms of call blocking and dropping probabilities. EFGC can readily
support multiple classes of traffic by assigning a separate acceptance ratio to each class. However,
computing these acceptance ratios in order to satisfy the desired QoS is not trivial.
ACKNOWLEDGMENT
The authors would like to thank the anonymous reviewers for their valuable comments. This
research was supported in part by Nortel Networks and by the Natural Sciences and Engineering
Research Council (NSERC) of Canada.
REFERENCES
[1] U. Varshney and R. Jain, “Issues in emerging 4G wireless networks,”IEEE Computer, vol. 34, no. 6, pp. 94–96, June
2001.
[2] S. Y. Hui and K. H. Yeung, “Challenges in the migration to 4G mobile systems,”IEEE Computer, vol. 41, no. 12, pp.
54–59, Dec. 2003.
[3] D. Hong and S. S. Rappaport, “Traffic model and performance analysis for cellular mobile radio telephone systems with
prioritized and nonprioritized handoff procedures,”IEEE Trans. Veh. Technol., vol. 35, no. 3, pp. 77–92, Aug. 1986, see
also: CEAS Tech. Rep. No. 773, College of Engineering and Applied Sciences, State University of New York, June 1999.
[4] M. Naghshineh and M. Schwartz, “Distributed call admission control in mobile/wireless networks,”IEEE J. Select. Areas
Commun., vol. 14, no. 4, pp. 711–717, May 1996.
[5] D. Levine, I. Akyildiz, and M. Naghshineh, “A resource estimation and call admission algorithm for wireless multimedia
networks using the shadow cluster concept,”IEEE/ACM Trans. Networking, vol. 5, no. 1, pp. 1–12, Feb. 1997.
[6] S. Choi and K. G. Shin, “Predictive and adaptive bandwidth reservation for handoffs in QoS-sensitive cellular networks,”
in Proc. ACM SIGCOMM’98, vol. 27, Vancouver, Canada, Oct. 1998, pp. 155–166.
[7] B. M. Epstein and M. Schwartz, “Predictive QoS-based admission control for multiclass traffic in cellular wireless
networks,” IEEE J. Select. Areas Commun., vol. 18, no. 3, pp. 523–534, Mar. 2000.
31
[8] A. Aljadhai and T. F. Znati, “Predictive mobility support for QoS provisioning in mobile wireless networks,”IEEE J.
Select. Areas Commun., vol. 19, no. 10, pp. 1915–1930, Oct. 2001.
[9] R. Ramjee, D. Towsley, and R. Nagarajan, “On optimal call admission control in cellular networks,”ACM/Kluwer Wireless
Networks, vol. 3, no. 1, pp. 29–41, Mar. 1997.
[10] S. Wu, K. Y. M. Wong, and B. Li, “A dynamic call admission policy with precision QoS guarantee using stochastic control
for mobile wireless networks,”IEEE/ACM Trans. Networking, vol. 10, no. 2, pp. 257–271, Apr. 2002.
[11] B. Li, L. Yin, K. Y. M. Wong, and S. Wu, “An efficient and adaptive bandwidth allocation scheme for mobile wireless
networks using an on-line local estimation technique,”ACM/Kluwer Wireless Networks, vol. 7, no. 2, pp. 107–116, 2001.
[12] B. Epstein and M. Schwartz, “Reservation strategies for multi-media traffic in a wireless environment,” inProc. IEEE
VTC’95, vol. 1, Chicago, USA, July 1995, pp. 165–169.
[13] J. E. Wieselthier and A. Ephremides, “Fixed- and movable-boundary channel-access schemes for integrated voice/data
wireless networks,”IEEE Trans. Commun., vol. 43, no. 1, pp. 64–74, Jan. 1995.
[14] M. C. Young and Y.-R. Haung, “Bandwidth assignment paradigms for broadband integrated voice/data networks,”Elsevier
Computer Communications Journal, vol. 21, no. 3, pp. 243–253, 1998.
[15] H.-H. Liu, J.-L. C. Wu, and W.-C. Hsieh, “Delay analysis of integrated voice and data service for GPRS,”IEEE Commun.
Lett., vol. 6, no. 8, pp. 319–321, Aug. 2002.
[16] D.-S. Lee and C.-C. Chen, “QoS of data traffic with voice handoffs in a PCS network,” inProc. IEEE GLOBECOM’02,
vol. 2, Taipei, Taiwan, Nov. 2002, pp. 1534–1538.
[17] M. A. Marsan, P. Laface, and M. Meo, “Packet delay analysis in GPRS systems,” inProc. IEEE INFOCOM’03, vol. 2,
San Francisco, USA, Mar. 2003, pp. 970–978.
[18] Y.-R. Haung, Y.-B. Lin, and J.-M. Ho, “Performance analysis for voice/data integration on a finite-buffer mobile system,”
IEEE Trans. Veh. Technol., vol. 49, no. 2, pp. 367–378, Mar. 2000.
[19] L. Yin, B. Li, Z. Zhang, and Y.-B. Lin, “Performance analysis of a dual-threshold reservation (DTR) scheme for voice/data
integrated mobile wireless networks,” inProc. IEEE WCNC’00, vol. 1, Chicago, USA, Sept. 2000, pp. 258–262.
[20] H. Wu, L. Li, B. Li, L. Yin, I. Chlamtac, and B. Li, “On handoff performance for an integrated voice/data cellular system,”
in Proc. IEEE PIMRC’02, vol. 5, Lisboa, Portugal, Sept. 2002, pp. 2180–2184.
[21] C.-J. Chang, T.-T. Su, and Y.-Y. Chiang, “Analysis of a cutoff priority cellular radio system with finite queueing and
reneging/dropping,”IEEE/ACM Trans. Networking, vol. 2, no. 2, pp. 166–175, Apr. 1994.
[22] B. Li, L. Li, B. Li, and X.-R. Cao, “On handoff performance for an integrated voice/data cellular system,”ACM/Kluwer
Wireless Networks, vol. 9, no. 4, pp. 393–402, July 2003.
[23] R. Guerin, H. Ahmadi, and M. Naghshineh, “Equivalent capacity and its application to bandwidth allocation in high-speed
networks,” IEEE J. Select. Areas Commun., vol. 9, no. 7, pp. 968–981, 1991.
[24] F. P. Kelly, “Notes on effective bandwidths,” inStochastic Networks: Theory and Applications, F. P. Kelly, S. Zachary, and
I. Ziedins, Eds. Oxford, UK: Oxford University Press, 1996, pp. 141–168.
[25] M. Schwartz,Broadband Integrated Networks. New Jersey, USA: Prentice Hall, 1996.
[26] Y. Fang, I. Chlamtac, and Y.-B. Lin, “Channel occupancy times and handoff rate for mobile computing and PCS networks,”
IEEE Trans. Comput., vol. 47, no. 6, pp. 679–692, June 1998.
[27] Y. Fang and I. Chlamtac, “Analytical generalized results for handoff probability in wireless networks,”IEEE Trans.
Commun., vol. 50, no. 3, pp. 396–399, Mar. 2002.
32
[28] W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery,Numerical Recipes in C: The Art of Scientific Computing,
2nd ed. Cambridge, UK: Cambridge University Press, 1992.
[29] P. Whittle,Prediction and Regulation by Linear Least-Square Methods, 2nd ed. Minnesota, USA: University of Minnesota
Press, 1983.
[30] P. J. Brockwell and R. A. Davis,Time Series: Theory and Methods, 2nd ed. New York, USA: Springer-Verlag, 1991.
[31] C. Jedrzycki and V. C. M. Leung, “Probability distribution of channel holding time in cellular telephone systems,” inProc.
IEEE VTC’96, vol. 1, Atlanta, GA, May 1996, pp. 247–251.
[32] M. M. Zonoozi and P. Dassanayake, “User mobility modeling and characterization of mobility patterns,”IEEE J. Select.
Areas Commun., vol. 15, no. 7, pp. 1239–1252, Sept. 1997.
[33] R. Guerin, “Channel occupancy time distribution in a cellular radio system,”IEEE Trans. Veh. Technol., vol. 35, no. 3,
pp. 89–99, 1987.
Majid Ghaderi received his B.S. and M.S. degrees in computer engineering from Sharif University
of Technology, Iran, in 1999 and 2001, respectively. He is currently a Ph.D. student in the School of
Computer Science at the University of Waterloo, Canada. His research interests are in the area of mobile
communication systems including radio resource management, admission control, and quality of service.
Raouf Boutaba is currently an Associate Professor in the School of Computer Science of the University
of Waterloo. Before that he was with the Department of Electrical and Computer Engineering of the
University of Toronto. Before joining academia, he founded and was the director of the telecommunications
and distributed systems division of the Computer Science Research Institute of Montreal (CRIM). Dr.
Boutaba conducts research in the areas of network and distributed systems management and resource
management in multimedia wired and wireless networks. He has published more than 120 papers in
refereed journals and conference proceedings. He is the recipient of the Premier’s Research Excellence Award, a fellow of the
faculty of mathematics of the University of Waterloo, and a distinguished lecturer of the IEEE Computer Society. Dr. Boutaba
is the Chairman of the IFIP Working Group on Networks and Distributed Systems, the Vice Chair of the IEEE Communications
Society Technical Committee on Information Infrastructure, and the Chair of the IEEE Communications Society Committee on
Standards. He is the founder and editor in Chief of the IEEE ComSoc eTransactions on Network and Service Management,
on the advisory editorial board of the Journal of Network and Systems Management, on the editorial board of the KIKS/IEEE
Journal of Communications and Networks, and the editorial board of the Journal of Computer Networks.